Higher-Frequency Analysis of Low-Frequency Data

Abstract

Low-frequency data present significant challenges to investors because infrequent observations lead to imperfect intra-period understandings of market behavior, which in turn frustrate risk management and portfolio monitoring. Nowhere is this phenomenon more vexing than in portfolios that include both liquid and illiquid investments, whose low-frequency reporting compromises an investor’s ability to form a holistic portfolio perspective. The authors examine the consequences of this issue and evaluate two methodologies for interpolating low-frequency data to facilitate use in higher-frequency settings. Through a numerical study, they determine that the decision to use one methodology over another—or whether to interpolate at all—is context-dependent, and illustrate the effect of higher-frequency interpolation on risk evaluation, attribution, and other common risk measures.